Huh? The linked article is nothing more than "this guy is black, so therefore helping any underprivileged black people gain university admissions is bad"
It's outrageous racism. A conclusion about all minorities based on one person's math mistake, where the logic is entirely based on shared skin color.
If you replace the races and make it a conclusion about legacy admissions or something, it's obviously stupid and illogical, right?
"This white guy doesn't know Afghanistan from Kazakhstan. More proof legacy admissions is bad!"
There's a bunch of needlessly inflammatory bullshit in that article. "Innumerate woke Bolshevik" and making fun of someone because he thinks she looks like a Harry Potter character. This guy seems like nothing more than a high school bully. E-mailing someone asking them to respond is nothing more than a fig leaf.
> Anyone who signed that petition is not only my personal enemy, but the enemy of free speech, the enemy of the spirit of the academy, and the enemy of western civilization.
As a side and probably fully off topic note (although...), I asked ChatGPT an innocent code question while not giving my code. It basically answered with the variable name I had written in my own code (da.ma.st) (a variable inside an object inside an object : data.main.stance). I still have to understand how and why it happened (I am not using anything else than ChatGPT in my browser and I absolutely never provided this chunk of code to the AI).
I further noticed that while I had a chatgpt window open, my dev site window was becoming laggy after many refreshes as if something was deliberately trying to scan it every time it got refreshed. I suspect the AI to scan other open tabs and simply reading through everything it can encounter. It is actually the only explanation up to date (but I unfortunately don't have much time to try to validate this speculative opinion: I will surely give other shots in order to narrow my suspicions).
I tend to think that this kind of data extraction frenziness may be a big problem in the future. Read it as : "let's collect everything, just tell everyone we are not collecting it, and then we'll see what we can get from it." Imagine such data being used in the future versions of governments if things get wild.
In the past this was a major issue that meant useful features were only ever usable after IE/Safari finally supported them half a decade later, but it has seriously gotten better. Sadly as a result of Chromium's overbearing presence, but it's a helpful outcome at least.
Problem with safari though is that it’s tied to OS updates that many people just defer for insanely long periods of time. So unlike the other browsers, it’s not evergreen, so if you need to support any iOS users or Mac users who don’t use chrome etc, you’re out of luck
Yeah that definitely sucks. I have seen MacOS Safari updates come though separately in Preferences.app > General > Software Update, but I think that release channel is for security issues.
Saying that, MacOS and iOS generally (up until recently, from what I've heard) have very good uptake rates for major updates. It's become less awful standards-wise as time has gone on in my experience at least.
That would actually fix some ugly CSS I have. The demo works. Neat.
Except... the demo doesn't use either the old syntax or the new syntax. The browser support is wrong (Firefox doesn't support it, the site says Firefox 16+; it says Chrome 43+ but in reality it's much newer: Chrome 148+). It says "Since 2018" but the spec was introduced in 2024.
So maybe an interesting overview of things that might be available or might not, but the filtering and data on the site doesn't seem to be useful.
Firefox is pretty irrelevant nowadays. They've dragged their feet for years when it comes to implementing new stuff, and now web devs don't even bother checking Firefox. Because devs know it won't work on ancient browsers, no need to confirm.
My personal trigger events were when Firefox didn't optimize DataView for the longest time, initially refused to implement import maps, and couldn't get WebGPU support done. At that point I lost interest in supporting it.
"widely available" has a precise meaning that includes Firefox (both desktop and Android). it might be irrelevant for some, but let's not twist industry definitions
This website says certain features work on firefox. But they don't. You can disregard firefox if you like. But if this "Modern CSS Code Snippets" website explicitly tells me their snippets work in firefox, I expect the snippets to work in firefox. Many of them do not.
again, "widely available" should not be intended in the general sense but as a much more precise industry term. "Baseline widely available" is defined[1] as a feature which has been available on all the core browsers (Chrome desktop and Android, Edge, Firefox desktop and Android, Safari on Mac and iOS) for two and a half years
I don't really care about someones phony definition of widely available. If it runs on 90% of user's browsers, it's widely available. I'll gladly make a web page that puts this definition online so that you can also reference it in discussions, if you want.
It's 2026, the most useful stuff was implemented over a decade ago. Stop trying to make the web platform do everything when it wasn't designed for that.
It doesn't matter what it was designed for 30 years ago. Computers also weren't designed to be put in your pocket, yet here we are. Things evolve, and browsers that do not keep up will eventually stop being used.
Firefox could (should?) be better in several aspects but it seems excessive to say it is pretty irrelevant.
It has 4.5% market share in Europe, 9% in Germany (statcounter numbers).
It is the browser that got the Google Labs folks to write a Rust jxl decoder for it, and now, thanks in part to that, Chrome is re-adding support for jxl.
You can be unhappy with Firefox (I often am myself), and Firefox HAS lost relevance, but can you really say it has become pretty irrelevant?
Also one of the features that I want -- scrollbar-gutter: stable -- is shown everywhere as being stable for many Safari versions, but when I try it it just didn't work. ¯\_(ツ)_/¯
One could argue with its stance, but I took it as a given (the equation for cognitive debt touches on science).
It feels entirely logical to view LLMs/coding agents as an almost final step in the short-term focus the overall system has been thriving on.
reply